Cervical cancer kills more than 200,000 people each year worldwide, disproportionately affecting women of low socioeconomic status. Infection with high-risk human papillomavirus (HPV) is the main causal factor, but additional factors must be involved because only a small proportion of HPV-infected women develop cancer. Epidemiologic studies suggest that some predisposing factors are genetic heritability accounts for about 27 percent of the total variation in liability to cervical tumor development. Because the responsible genes (or gene variants) have not been uncovered, it is not yet possible to develop methods for identifying the small proportion of women with preinvasive cervical intraepithelial neoplasia (CIN) who will need treatment. To address this deficit, we plan to identify markers that associate with cervical cancer. Such markers might also be useful for screening women in the general population. To elucidate the role of genetic factors in the development of cervical cancer, we will study DNA from women who have invasive cervical cancer (ICC) or CIN III and are also infected with high risk HPV subtypes. We will select candidate genes that appear critical for the development of the cancer, including the HLA DQB1/DRB1 locus and other immune markers, genes that are cellular targets of HPV E6, E7, and E5 oncoproteins, and genes implicated in the progression of cervical neoplasia. Within each gene, we will focus on small variations, such as differences in single bases. Such single nucleotide polymorphisms (SNPs) are the most common genetic variations among individuals, accounting for a substantial proportion of phenotypic variability. We will evaluate their influence on interactions between HPV and cervical cells by using the transmission/disequilibrium test (TDT). Unlike the case-control method, the TDT assesses associations between specific alleles and disease endpoints without being vulnerable to errors caused by stratification of genetically disparate populations or undiagnosed preclinical disease in "controls.'" We will test each variation for association with ICC or CIN Ill by comparing frequency distributions of patient alleles (cases) with those of nontransmitted parental 'alleles, which provide a perfect ethnically matched control sample. If a polymorphism is inherited with higher-than-Mendelian frequency by the women with ICC or CIN III, we will suspect that variant of predisposing HPV-infected women to cervical cancer. Subsequent case control and cohort studies could then confirm the association. This investigative sequence is much more cost-effective than population studies that begin without first identifying potentially culpable genes. Determining how small variations in host DNA influence vulnerability to cervical cancer is critical to understanding the pathogenesis of the disease and therefore to the development of superior screening and diagnostic tests. As the field of pharmacogenetics expands, this information might enable vaccine and drug developers to tailor their products to "cervical cancer genotypes."